psu's

[odysseybmx414]

ok, so ive been testing different psu's to find out which is the better one i tried a ct trafo with 1 10000uF on each rail and a snubber(res+101cap)with that. sounded great then i took off the snubbers and it still sounded ok. then i tried something different. i put one of the 10000uF caps +side to +rail, and -side to -rail, but i kept the snubbers going from each rail to ground. each of these psus were tested on a opa541 noninverting amp. this sounded better than all the others it sounded like it had more "punch" and clearer sound and also, sounded more detailed.

just thought i would put this out there so people could think about it

P.S. the opa541 was bypassed with 2 220uF nichicon caps

[jaudio]

then i tried something different. i put one of the 10000uF caps +side to +rail, and -side to -rail,

Im a little confused are the caps connected as they are normally?

but i kept the snubbers going from each rail to ground. each of these psus were tested on a opa541 noninverting amp. this sounded better than all the others it sounded like it had more "punch" and clearer sound and also, sounded more detailed.

Where the snubbers kept unconnected?

[odysseybmx414]

-----------+CAP---------V-


-----------GND----------


------------CAP(-)-------V+


the positive terminal of the 10000uF cap is connected to V+
the negative terminal of the cap is at V-

the snubbers still go from each rail to ground

so there is only 1 10000uF cap going from positive rail to negative rail(no connection to ground)

this was used on opa541 and sounded better than all other configs (IMO)

just thot i would put this out there so people could try it and see if they like it as much as i do.


another thing when there are 2 10000uF caps (going from each rail to ground) there is only 5000uF of capacitance between the rails because it divides the capacitance like putting resistors in parallel. this way you can use only 1 capacitor and get twice the capacitance. IMO this sounded more controlled and detailed and just better than the others.

[jaudio]

Thank alot that one more thing to look into

[john65b]

I don't know if this is correct, but if you previously had your power cap + to + rail and + to - rail, then you had them connected wrong in the first place and that explains why it sounds better now.

Consider yourself lucky you didn't blow up your - rail cap...

[john65b]

Whoops, I think I know what you are talking about now. Sorry...I can try it later...

[odysseybmx414]

lol glad u figured it out yes try it

[odysseybmx414]

i just noticed that in the little diagram thing i have the + side of cap to - rail thats not right it should be - side of cap to negative rail and positive sid of cap to positive rail sorry guys stupid mistake

[butler853]

I think I remember Nelson Pass commenting about doing this, but if memory serves me correctly, he also suggested using caps that were rated at 4X the rail voltage.

After a half hour of searching I found a somewhat relevant post.

http://www.diyaudio.com/forums/showt...431#post516431

Maybe I confused 4X capacitance with the required voltage rating. Looking at the power supply schematic, they are using 120V rails and 250V caps. So maybe it's twice the rail voltage. Hopefully someone more knowledgeable than me can chime in. But my point is that you might be exceeding the voltage rating of your capacitors.

[jaudio]

OK so where the caps would be normally connected to ground, they are connected to each other?


---------------(+)-----------pos rail
c
a
p
---------------(-)
connected
---------------(+)
c
a
p
--------------(-)-------------- neg rail